simple schematics AND ASM-1!!!!

[Magnus Danielson]

From: "Roel Das" <Roel.Das at student.groept.be>
Subject: simple schematics
Date: Sun, 9 Jan 2000 15:21:41 +0100

> I'd like to have some simple noise sources. I'm in this project i which we
> have to play a live soundtrack over a very old silent movie ('He who gets
> slapped', 1920 or something? I don't remember). I'd like to build some
> simple modules to add some random noises in between melodic pieces. They
> don't have to be controllable, just very simple. And maybe a very basic
> mixer to get them on one channel. I don't have alot of time to build them
> (from feb 5 to feb 20).

I just uploaded the crisply fresh schematic of the ASM-1 Noise Source.
It is a simple design taking not an awfull amount of components.

Some excuses for the schematic, it is somewhat of a quick-job since the right
component was not available and I just wanted something out. It's done in the
gEDA tool (http://www.geda.seul.org/) straigth out of the development tree ;)

Anyway, if you don't need the pink noise and random noise outputs you can just
drop that op-amp and components out of it. The schematic looks like it eats
4 ICs, but it actually only takes 2 ICs since the TL082s are duals.

I just realized that I missed to set the transistor, but it should be an
2N3904 or similar.

I hope this curcuit is near what you are looking for.

> And i've been playing with an idea. What sound do you get when you use a
> pulse to clock a pseudo-random generator (shift register with an exor or
> something like that)? And maybe connect this to an up down counter which you
> feed on the trigger to get symmetry?

You can use a shift register with xor gates to get noise, but you need to
clock it fairly high and use a pretty long shift register (at least 16 bits
but preferably longer) to get good noise. To low frequency will cause the
upper end to lack good noise energy. To short register causes a most annoying
"mechanical" pulsing so you would not want that. Also, having a too short
shift register will also make the number of frequencies too few, so if you
would at any given point have a have a high-Q filter or narrow BP filter only
a few frequencies would come out and not a narrow range of frequencies.
A longer register will effectively create more frequencies evenly spaced out
and given enougth length will they become dense enougth for most practical
usage approximate white noise.

For a fair treatment of noise sources, I recommend a visit into "The Art of
Electronics" by our friends Horwitz and Hill. Their example curcuit in Fig 9.90
sould be easy enougth to build while providing a usefull addition to most
people's measurement rigs.

Either way, it will be more components than the ASM-1 noise source has.

I think that the ASM-1 solution should be easy enougth for you to have up and
working in basically no time. Notice that the power is sent trougth filtering
before reaching the op-amps. I think the 100k from the transistor should also
see that filtered power...

Cheers,
Magnus